007) 8424–8429www.elsevier.com/locate/tsf

Thin Solid Films 515 (2

Direct imaging by atomic force microscopy of surface-localizedself-assembled monolayers on a cuprate superconductor and surface X-ray

scattering analysis of analogous monolayers on the surface of water

Steen B. Schougaard a,b,⁎, Niels Reitzel c, Thomas Bjørnholm c, Kristian Kjaer d,e,Torben R. Jensen h, Olga E. Shmakova f, Ramon Colorado Jr. f, T. Randall Lee f,

J.-H. Choi g, John T. Markert g, David Derro g, Alex de Lozanne g, John T. McDevitt b

a Département de Chimie, Université du Québec à Montréal, Case postale 8888, Succ. Centre-ville, Montréal, Québec, Canada H3C 3P8b Texas Materials Institute, Center for Nano and Molecular Science and Engineering, Department Chemistry and Biochemistry,

The University of Texas at Austin, Austin, TX, 78722, USAc Nano-Science Center, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen, Denmarkd Max-Planck Institute of Colloids and Interfaces, Am Mühlenberg, D-14476 Potsdam/Golm, Germany

e Niels Bohr Institute, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen, Denmarkf Department of Chemistry, University of Houston, Houston, TX 77204-5003, USA

g Department of Physics, The University of Texas at Austin, Austin, TX 78712-1081, USAh Interdisciplinary Nanoscience Center (iNANO), Department of Chemistry, University of Aarhus, DK-8000 Aarhus C, Denmark

Received 22 March 2007; accepted 5 April 2007Available online 18 April 2007

Abstract

A self-assembled monolayer of CF3(CF2)3(CH2)11NH2 atop the (001) surface of the high-temperature superconductor YBa2Cu3O7-x wasimaged by atomic force microscopy (AFM). The AFM images provide direct 2D-structural evidence for the epitaxial 5.5 Å square √2×√2R45°unit cell previously predicted for alkyl amines by molecular modeling [J.E. Ritchie, C.A. Wells, J.-P. Zhou, J. Zhao, J.T. McDevitt, C.R.Ankrum, L. Jean, D.R. Kanis, J. Am. Chem. Soc. 120 (1998) 2733]. Additionally, the 3D structure of an analogous Langmuir monolayer ofCF3(CF2)9(CH2)11NH2 on water was studied by grazing-incidence X-ray diffraction and specular X-ray reflectivity. Structural differencesand similarities between the water-supported and superconductor-localized monolayers are discussed.© 2007 Elsevier B.V. All rights reserved.

Keywords: Atomic force microscopy; Surface composition; Grazing-incidence X-ray diffraction (GIXD); Self-assembled monolayers (SAMs)

1. Introduction

The process of forming self-assembled monolayers (SAMs)on copper oxide-based high-temperature superconductors isknown to remove corrosion products initially and to formdensely packed protective layers subsequently [1,2]. SAMs areformed when ceramic or thin-film samples of various copperoxide superconductor materials are exposed to organic or metal

⁎ Corresponding author. Département de Chimie, Université du Québec àMontréal, Case postale 8888, Succ. Centre-ville, Montréal (Québec), CanadaH3C 3P8.

E-mail address: schougaard.steen@uqam.ca (S.B. Schougaard).

0040-6090/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.tsf.2007.04.034

organic molecules terminated with an alkylamine group [3–5].Their structure is of key interest since formation of SAMs offersa convenient way to produce inorganic/organic compositestructures with well-defined interfaces. This methodology hasbeen successfully exploited with normal conductors, semicon-ductors, and insulators [6–8]. Here, we present the first directexperimental structural evidence for the formation of highlyordered SAMs of CF3(CF2)3(CH2)11NH2 atop a (001) surface ofthe high-temperature superconductor YBa2Cu3O7−x, as imagedby deflection mode atomic force microscopy (AFM).

The adsorbates employed here differ from those previouslystudied using computer modeling [1] and reflectance angleinfrared spectroscopy [1] since the alkyl chains of the amines are

8425S.B. Schougaard et al. / Thin Solid Films 515 (2007) 8424–8429

partially fluorinated to enhance crystallinity and to provide amore robust protective overlayer for the superconductor material[9].We used two different partially fluorinated adsorbates becausepreliminary atomic force microscopy (AFM) studies found thatonly the shorter molecule (CF3(CF2)3(CH2)11NH2) gave SAMsthat showed an ordered registry on the surface of the super-conductor; conversely, Langmuir isotherm measurements foundthat that only the longer molecule (CF3(CF2)9(CH2)11NH2) gavedensely packed Langmuir monolayers on water. Structuralcharacterization of the latter system by both grazing-incidenceX-ray diffraction (GIXD) and specular X-ray reflectivity (XR)was needed to establish the structure and order of amine-basedpartially fluorinated monolayer films, which are reported herefor the first time. In particular, we wished to determinewhether the structure and packing of the films were influencedby epitaxial registry between the adsorbates and the surface ofthe superconductor.

2. Experimental details

2.1. Synthesis

The synthesis of the fluorinated amines (CF3(CF2)n-(CH2)11NH2, n=3 or 9) has been described previously [10].

2.2. Crystal growth and mounting

The crystal growth and mounting process is similar to theone previously described in Edwards et al. [11]. Briefly,crystals are grown by slow cooling from 1000 °C in a bariumcopper oxide flux. They are mechanically removed andannealed at 450 °C in flowing oxygen for several weeks andslowly cooled. The resulting thin plate-like crystals (approxi-mately, 1 mm×1 mm×20 μm) of YBa2Cu3O7−x (x≈0.05)have the c axis perpendicular to the largest surfaces. Crystalswere mounted with silver epoxy between two copper wirestuds. Excess epoxy was removed using a file before cleaving.

2.3. SAM formation process

YBa2Cu3O7−x crystals are cleaved at room temperature in airimmediately before they are soaked in a 1 mM CF3(CF2)3-(CH2)11NH2/hexane solution (high pressure liquid chromato-graphy grade) for 24 h. The crystals are removed from thesolution, thoroughly rinsed with pure hexane, and blown drywith N2 before transfer to the AFM (Nanoscope III, DigitalInstruments). Images were collected in deflection mode, using astandard Si3N4 tip (DN-type, Digital Instruments). Care wastaken in minimizing the z-axis feedback loop gain whilemapping the cantilever deflection.

2.4. Grazing-incidence X-ray diffraction (GIXD) and specularX-ray reflectivity (XR)

A monolayer of CF3(CF2)9(CH2)11NH2 was compressed to asurface area per molecule of ∼30±1 Å2 at a surface pressure ofΠ ∼35 mN/m on a pH=11 subphase at 21.3±0.1 °C. Data

were collected using the liquid surface diffractometer [12] at theBW1 undulator beamline [13] at the synchrotron radiationfacility HASYLAB, DESY, Hamburg, Germany, using awavelength of λ=1.304 Å monochromated by a Be crystal[14,15]. Two different X-ray techniques were used [12,16,17]:GIXD and XR.

For GIXD, the monolayer was X-ray illuminated at a grazingangle of incidence (αi) slightly below the critical angle (αc)for total reflection (αi =0.85αc), thus increasing the surfacesensitivity by minimizing the penetration depth of the incidentX-rays into the water subphase. Any lateral crystallinity in themonolayer will then give rise to Bragg rods [12,16,17].The scattered X-ray intensity (I) was measured vs. horizontalscattering angle (2θxy) and vertical exit angle (αf) and convertedto I(Qxy,Qz) where the vertical and horizontal scattering vec-tor components are Qz≅ (2π/λ)sin(αf) and Qxy≡ (Qx

2 +Qy2)1/2≅

(2π/λ)[1+cos2(αf)−2cos(αf)cos(2θxy)]1/2 [12,16]. Note that theLangmuir monolayer is a 2D powder: 2D-crystallites occur thatall have their base planes horizontal but represent all azimuthalorientations, so that only Qxy≡ (Qx

2 +Qy2)1/2 can be resolved, not

Qx, Qy individually.The XR experiments probe the laterally averaged electron

density profile (ρ(z)) normal to the interface by varying theincident and exit angles (αi,αf) simultaneously (αf =αi≡α),recording the intensity pattern resulting from interferencebetween rays reflected at different depths. Here both laterallycrystalline and non-crystalline parts contribute. The mea-sured reflectivity, R(Qz), normalized to the Fresnel reflec-tivity, RF(Qz), calculated for a theoretical sharp air/waterinterface, is shown vs. the purely vertical scattering vector Qz=(2π/λ)[sin(αi) + sin(αf)] = (4π/λ)sin(α). The XR data wereinverted to yield ρ(z) by a method similar to that described inPedersen and Hamley [18].

3. Results and discussion

One of the significant challenges associated with the directimaging of cuprate superconductor-localized monolayers hasbeen identifying appropriate sample specimens with surfaceroughness sufficiently low for these measurements. In the morecommonly exploited SAM systems, such as thiols on gold, lowroughness can be attained by simply polishing and subsequentlyannealing the gold surface at elevated temperatures or byevaporating the metal onto cleaved mica [19,20]. Similarprotocols cannot be used with the YBa2Cu3O7−x systembecause of its highly reactive nature and complex composition[21]. While previous studies have shown bulk ceramic samplesand laser-ablated cuprate films to be suitable for localizingamine monolayers, these samples have proven too rough forstructural characterization of the monolayers by AFM.

To overcome these limitations, a procedure for makingSAMs using cleaved YBa2Cu3O7−x single crystals has beendeveloped. Here, crystals of YBa2Cu3O7−x (x≈0.05) aremounted and cleaved using an adaptation of a previouslyreported method, whereby the crystals are fragmented at low-temperature for use in STM imaging studies [11]. These studiesshowed that the cleavage of such crystals occurs primarily

Fig. 1. The crystal structure of YBa2Cu3O7−x , including the atom labelingscheme (A) for (B) and (C). The box indicates the contents of one unit cell. Thestructure before (B) and after (C) cleavage. The arrow indicates the cleavageplane between the Cu–O chain layer and the Ba–O sheet layer.

Fig. 3. (A) The proposed superstructure of the amines on the copper oxide chainlayer. Dotted line: YBa2Cu3O7−x unit cell. Solid line: √2×√2R45° SAM-superstructure. Note that only every other electrophilic copper site is occupied.(B) Space filling model of the molecule used for the AFM study. The molecularfootprint is shown as cross-hatched circles.

Fig. 2. Direct deflection mode AFM images (30×30 nm) of CF3(CF2)3(CH2)11NH2

on YBa2Cu3O7−x (A). A Fourier-transform of image (A) is shown in (B). Arrowsmark the peaks that reflect the square 5.6 Å unit cell. (C) HOPG reference scan(10×10 nm) obtained using the same tip after image Awas collected.

8426 S.B. Schougaard et al. / Thin Solid Films 515 (2007) 8424–8429

between the Cu–O chain layer and the Ba–O sheet layer (Fig. 1)[11].

Initial AFM scans of large areas at low resolution of theSAM atop the cleaved crystal show that the surface roughnessis comparable to the highly ordered pyrolytic graphite (HOPG)reference. At higher resolution, atomically resolved AFMimages were obtained (Fig. 2A), consistent with a 5.6±0.5 Åsquare unit cell of area ∼31 Å2. This unit cell does notcorrelate with any known surface reconstruction of pristineYBa2Cu3O7−x [11,22,23], and the 5.6 Å spacing is similar inmagnitude to lattice spacings measured for related SAMs ongold derived from partially fluorinated alkanethiols [24]. Theunit cell was reproducibly observed in different areas on thecrystal surface using various scan areas, different scandirections, and a variety of scan speeds [25]. Examination ofthe Fourier-transformed image (Fig. 2B) reveals that the peaksare weaker along one reciprocal axis than the other. Thisphenomenon arises from the striped nature of the direct image,the origin of which is presently unclear. However, it is certainthat the stripes are not an artifact of the AFM tip, asdemonstrated by subsequent scans of the HOPG reference(Fig. 2C).

Highly ordered SAM structures are believed to form pri-marily due to the interplay between two driving forces [27]:(1) the interaction of the headgroup with the substrate surface,and (2) the lateral intermolecular tail-to-tail interactions. Forthe specific case of alkylamines adsorbed onto the cupratesuperconductor, the two types of interactions are between theamine headgroup and the YBa2Cu3O7−x surface, and betweenthe partially fluorinated alkyl chains. Thus, while the square5.6±0.5 Å repeat unit (of area 31 Å2) appears to be, withinerror, consistent with a √2×√2R45° unit supercell (of area29.8 Å2) of the (001) surface [28] (Fig. 3A), it could simply becaused by a maximization of the tailgroup interactions.Therefore, to determine whether the unit cell arises from theregistry of the superconductor surface epitaxially, similar

monolayers were prepared on water surfaces using theLangmuir method. Having no preferred bonding sites, thesurface of liquid water is preferred over solids, such as Au or

Fig. 5. Grazing-incidence X-ray diffraction data of the CF3(CF2)9(CH2)11NH2

Langmuir monolayer. (A) Plot of the X-ray intensity (grey scale) vs. Qxy and Qz,the horizontal and vertical components of the scattering vector [12,16]. (Solidline: The peak of the observed 5.78±0.10 Å hexagonal unit cell. Dashed lines:the hypothetical reflections of a square 5.6 Å unit cell as found in the monolayer

8427S.B. Schougaard et al. / Thin Solid Films 515 (2007) 8424–8429

Ag. The Langmuir monolayer is therefore primarily defined bythe tail-to-tail interactions.

The structural features of the water-localized film com-pressed to a dense monolayer (Fig. 4) were resolved by meansof grazing-incidence X-ray diffraction (GIXD) [12,16,17]. Thealkylamine used for this Langmuir film was longer by sixCF2 units than the alkylamine used on YBa2Cu3O7−x. Thissubstitution was motivated by both experimental and theoreticalstudies showing that the replacement of flexible hydrocarbonsby stiffer fluorocarbons increases the crystallinity of Langmuirfilms [29,30], which, together with the higher scattering powerof fluorine over hydrogen, should enhance the X-ray signal.

GIXD data are shown in Fig. 5 as intensity (grey scale) vs.Qxy and Qz, the horizontal and vertical components of thescattering vector [12,16]. For a Langmuir film consisting of 2Dcrystallites all with their base planes horizontal, the Bragg peakat Qxy=1.255±0.02 Å−1 would extend along Qz as a ‘Braggrod’ (at constant Qxy) to ca. Qz≤π/(length of molecule) [16]. Inthe present data, the observed intensity extends along theScherrer ring of constant Qtotal

2 =Qxy2 +Qz

2 (full line). Thisobservation is indicative of a mosaic distribution of the orien-tation of the base planes of the monolayer domains, cf. Fig. 9g,hin ref. [31] and Fig. 2a,b in ref. [32]. Specular X-ray reflectivitydata (Fig. 6A) were inverted to yield the laterally averagedelectron density distribution ρ(z) across the water/film/airinterface (Fig. 6B). Defining z=0 (the film/air interface),where ρ is half of its maximum value, a measure of the filmthickness, H, may be derived by integrating ρ(z)Amol fromabove the interface until, at z=−H (Fig. 6B), the N=346electrons of the molecule are accounted for:

NuAmol

Z þl

�HqðzÞdz

Using the area per molecule Amol=30±1 Å2 yields H=26±1 Å. The total length of the molecule should be about28.7±0.5 Å. However, from space filling considerations(vide infra), we expect the hydrocarbon moiety to be tiltedby ca. 48°±3°, giving an expected monolayer thickness of

Fig. 4. Compression isotherm of the Langmuir CF3(CF2)9(CH2)11NH2

monolayer. The arrow indicates the compression at which the X-ray data weregathered.

on a superconductor surface.) (B) Plotted vs. Qxy is the intensity integrated overthe Qz intervals indicated (and marked with brackets in A). The X-raywavelength was λ =1.304 Å.

23.6±0.6 Å, which is similar to the 23±1 Å value measured fora CF3(CF2)9(CH2)11SH SAM on Au [24]. The larger value ofH=26±1 Å deduced from Fig. 6B might be due to the mosaicdistribution inferred from the GIXD data. Indeed, attempts wereunsuccessful to fit the reflectivity data (Fig. 6A) with moredetailed models of a non-mosaic monolayer. However, semi-quantitatively, the electron density curve (Fig. 6B) seemsconsistent with the proposed model of a mosaic monolayer withthe amine near the water interface and the fluorinated moietynear the air interface.

Comparison between the GIXD measurements of theCF3(CF2)9(CH2)11NH2 Langmuir film and the predicted peaksof the square unit cell found on the YBCO crystal clearly shows

Fig. 6. (A) Specular X-ray reflectivity for the monolayer on water, R(Qz),normalized by the Fresnel reflectivity, RF(Qz), calculated for an ideal abruptinterface between air and bulk water. (B) Electron density profile ρ(z) on anabsolute scale, inverted from the data in (A). As discussed in the text, z=0 is theair–film interface, taken to be where ρ(z) is half is maximum value, and theelectron density above z=−H=−26±1 Å accounts for all of the electrons ofthe molecule CF3(CF2)9(CH2)11NH2.

8428 S.B. Schougaard et al. / Thin Solid Films 515 (2007) 8424–8429

no correlation (Fig. 5A). We thus conclude that the bondingsites of the crystal surface influence the structure of the SAM.Furthermore, when exploring the possible bonding sites for thenucleophilic amine headgroup of all possible (001) YBa2Cu3O7

−x surfaces, the only electrophilic sites that are commensuratewith the square 5.6 Å supercell are the copper sites (Fig. 3A)[28]. So, in accordance with previous studies [1,33], it appearsthat coordination of the amine lone pairs to Cu is responsible forthe epitaxy between the SAM and the superconductor.

The three-dimensional SAM structure model is based on aclose packed structure that can be accommodated within theAFM unit cell without violating known minimal volumemolecular geometry relations. The area (29±1 Å2) of theLangmuir film 2D unit cell (a=5.78±0.10 Å) (Fig. 5A)combined with the sharp increase in the compression/forcecurve below ∼36 Å2/molecule (Fig. 4) shows that the 31±3 Å2

unit cell found on the SAM-covered YBa2Cu3O7−x surface canaccommodate no more than one molecule. From previousstudies of perfluorinated alkanes, it is known that the molecularcross sectional area is ∼28 Å2 [34]. The perfluorinated part of

the chain must therefore be approximately perpendicular to thesuperconductor surface (Fig. 3B). Similarly, previous studieshave shown that hydrocarbon chains in close-packed mono-layers have a cross-sectional area of ∼19–21 Å2 [35]. Thus, toobtain a close-packed structure, the hydrogenated alkyl chainmust be tilted at an angle of approximately arccos(20 Å2/30 Å2)=48°±3° from vertical as illustrated in Fig. 3B. Thekink or bend proposed here between the perfluorinated andhydrocarbon sections is comparable to the ones found inprevious structure studies of SAMs derived from partiallyfluorinated thiols [24,36].

4. Conclusions

We have presented a simple method for generatingpassivated surfaces of YBa2Cu3O7−x sufficiently stable andsmooth for atomic scale characterization by AFM.We have thenused this technique to provide the first direct evidence for anorganized monolayer atop a copper oxide superconductor. Thestructural motif found is consistent with prior structural modelsconsisting of a √2×√2R45° superstructure of amines atop the(001) plane of YBa2Cu3O7−x. The discovery that simplefluorinated amines may form highly ordered and densemonolayers might lead to their use as a molecular blockinglayer with sub-nanometer dimensions or as a vehicle forproducing more complex organic/inorganic epitaxial structures.This scheme holds promise as it combines the design flexibilityinherent to organic/metal organic molecules with the unusualelectronic properties of the high-temperature superconductorswithout compromising the reactive nature of the latter.

Acknowledgments

J. T. McDevitt thanks the AFOSR, NSF (DMR-0211150),TCSUH, and the Welch Foundation for support. S. B.Schougaard thanks the Danish Research Agency for financialsupport. K. Kjaer thanks the Carlsberg Foundation for financialsupport. A. de Lozanne thanks the NSF (DMR-0308575) andthe Welch Foundation (F-1533) for support. The synchrotron X-ray studies benefited from access to beamline BW1 atHASYLAB at DESY, Hamburg, Germany under the IHP-Contract HPRI-CT-1999-00040/2001-00140 of the EuropeanCommission and from support by the Danish Natural ScienceResearch Council's DANSYNC program. T. R. Lee thanks theNSF (DMR-0447588), the Texas Center for Superconductivity,and the Welch Foundation (E-1320) for support. J. T. Markertalso thanks the Welch Foundation (F-1191) for support.Dr. D. W. Breiby, Dr. D. A. Vanden Bout, Dr. C. Wells andDr. J. Mauzeroll are gratefully acknowledged for discussions.

References

[1] J.E. Ritchie, C.A. Wells, J.-P. Zhou, J. Zhao, J.T. McDevitt, C.R. Ankrum,L. Jean, D.R. Kanis, J. Am. Chem. Soc. 120 (1998) 2733.

[2] J.E. Ritchie, W.R. Murray, K. Kershan, V. Diaz, L. Tran, J.T. McDevitt, J.Am. Chem. Soc. 121 (1999) 7447.

[3] F. Xu, K. Chen, R.D. Piner, C.A. Mirkin, J.E. Ritchie, J.T. McDevitt, M.O.Cannon, O. Michael, D.R. Kanis, Langmuir 14 (1998) 6505.

8429S.B. Schougaard et al. / Thin Solid Films 515 (2007) 8424–8429

[4] K. Chen, C.A. Mirkin, R.-K. Lo, J. Zhao, J.T. McDevitt, J. Am. Chem.Soc. 117 (1995) 6374.

[5] J.T. McDevitt, S.G. Haupt, C.E. Jones, in: A.J. Bard, I. Rubinstein(Eds.), Electroanalytical Chemistry, vol. 19, Marcel Dekker, New York,1996, p. 337.

[6] D.L. Allara, Biosens. Bioelectron. 10 (1995) 771.[7] A. Ulman, An Introduction to Ultrathin Organic Films from Langmuir–

Blodgett to Self-Assembly, Academic Press, London, 1991.[8] A. Ulman, Chem. Rev. 96 (1996) 1533.[9] H. Fukushima, S. Seki, T. Nishikawa, H. Takiguchi, K. Tamada, K. Abe, R.

Colorado Jr., M. Graupe, O.E. Shmakova, T.R. Lee, J. Phys. Chem. 104(2000) 7417.

[10] T.R. Lee, M. Graupe, U.S. Patent No. 6509100, 21 Jan. 2003.[11] H.L. Edwards, J.T. Markert, A.L. de Lozanne, J. Vac. Sci. Technol. B 12

(1994) 1886.[12] T.R. Jensen, K. Kjaer, in: D. Möbius, R. Miller (Eds.), Novel Methods to

Study Interfacial Layers, vol. 11, Elsevier, Amsterdam, 2001, p. 205.[13] R. Frahm, J. Weigelt, G. Meyer, G. Materlik, Rev. Sci. Instrum. 66 (1995)

1677.[14] J. Als-Nielsen, A.K. Freund, Rev. Sci. Instrum. 63 (1992) 1156.[15] J. Als-Nielsen, K. Kjaer, Nucl. Instrum. Methods Phys. Res., A 323 (1992)

686.[16] K. Kjaer, Physica, B 198 (1994) 100.[17] J. Als-Nielsen, D. Jacquemain, K. Kjaer, F. Leveiller, M. Lahav, L.

Leiserowitz, Phys. Rep. 246 (1994) 251.[18] J.S. Pedersen, I.W. Hamley, J. Appl. Cryst. 27 (1994) 36.[19] C.A. Alves, M.D. Porter, Langmuir 9 (1993) 3507.[20] T.M. Schultz, PhD thesis, Aarhus University, 1998.[21] J.P. Zhou, R.K. Lo, J.T. McDevitt, J. Talvacchio, M.G. Forrester, B.D.

Hutn, Q.X. Jia, D. Reagor, J. Mater. Res. 12 (1997) 2958.

[22] H. Behner, K. Rührnschopf, W. Rauch, G. Wedler, Appl. Surf. Sci. 68(1993) 179.

[23] H.P. Lang, H. Haefke, G. Leemann, H.J. Guntherodt, Physica, C 194(1992) 81.

[24] K. Tamada, T. Ishida, W. Knoll, H. Fukushima, R. Colorado Jr., M.Graupe, O.E. Shmakova, T.R. Lee, Langmuir 17 (2001) 1913.

[25] AFM images of the unmodified surface could not be obtained presumablydue to reaction of the unprotected YBa2Cu3O7−x surface with atmosphericwater and CO2 [21,26].

[26] S.B. Schougaard, M.F. Ali, J.T. McDevitt, Appl. Phys. Lett. 84 (2004)1144.

[27] A.J. Pertsi, M. Grunze, J. Chem. Phys. 106 (1997) 7343.[28] H. Casalta, P. Schleger, P. Harris, B. Lebech, N.H. Andersen, R. Liang, P.

Dosanjh, W.N. Hardy, Physica, C 258 (1996) 321.[29] M.K. Bernett, M.A. Zisman, J. Phys. Chem. 67 (1963) 1534.[30] D. Jacquemain, S.G. Wolf, F. Leveiller, F. Frolow, M. Eisenstein, M.

Lahav, L. Leiserowitz, J. Am. Chem. Soc. 114 (1992) 9983.[31] I. Weissbuch, M. Berfeld, W.G. Bouwman, K. Kjaer, J. Als-Nielsen, M.

Lahav, L. Leiserowitz, J. Am. Chem. Soc. 119 (1997) 933.[32] R. Popovitz-Biro, R. Edgar, I. Weissbuch, R. Lavie, S. Cohen, K. Kjaer, J.

Als-Nielsen, E. Wassermann, L. Leiserowitz, M. Lahav, Acta Polym. 49(1998) 626.

[33] J. Zhu, F. Xu, S.J. Schofer, C.A. Mirkin, J. Am. Chem. Soc. 119 (1997)235.

[34] M.Y. Li, A.A. Acero, Z.Q. Huang, S.A. Rice, Nature 367 (1994) 151.[35] I. Kuzmenko, V.M. Kaganer, L. Leiserowitz, Langmuir 14 (1998) 3882.[36] S. Frey, K. Heister, M. Zharnikov, M. Grunze, K. Tamada, R. Colorado Jr.,

M. Graupe, O.E. Shmakova, T.R. Lee, Isr. J. Chem. 40 (2000) 81.